tarekziade HF Staff
Publish MIC usage, slim fixture and reviewed Transformers expectations
1d300e4 verified | { | |
| "fixture": "huggingface-mic/bart-base-mic", | |
| "key": "c89a9e951b4fc3be51d5ed5cc386f5d048e5cd8b22fc8efc7aeb5137234c020d", | |
| "key_fields": { | |
| "architectures": [ | |
| "BartModel" | |
| ], | |
| "output_hidden_states": false, | |
| "return_dict": true, | |
| "chunk_size_feed_forward": 0, | |
| "is_encoder_decoder": true, | |
| "id2label": { | |
| "0": "LABEL_0", | |
| "1": "LABEL_1", | |
| "2": "LABEL_2" | |
| }, | |
| "label2id": { | |
| "LABEL_0": 0, | |
| "LABEL_1": 1, | |
| "LABEL_2": 2 | |
| }, | |
| "problem_type": null, | |
| "vocab_size": 50265, | |
| "max_position_embeddings": 1024, | |
| "encoder_layers": 2, | |
| "encoder_ffn_dim": 128, | |
| "encoder_attention_heads": 4, | |
| "decoder_layers": 2, | |
| "decoder_ffn_dim": 128, | |
| "decoder_attention_heads": 4, | |
| "encoder_layerdrop": 0.0, | |
| "decoder_layerdrop": 0.0, | |
| "activation_function": "gelu", | |
| "d_model": 64, | |
| "dropout": 0.1, | |
| "attention_dropout": 0.1, | |
| "activation_dropout": 0.1, | |
| "init_std": 0.02, | |
| "classifier_dropout": 0.0, | |
| "scale_embedding": false, | |
| "use_cache": true, | |
| "pad_token_id": 1, | |
| "bos_token_id": 0, | |
| "eos_token_id": 2, | |
| "decoder_start_token_id": 2, | |
| "forced_eos_token_id": 2, | |
| "is_decoder": false, | |
| "tie_word_embeddings": true, | |
| "add_bias_logits": false, | |
| "add_final_layer_norm": false, | |
| "classif_dropout": 0.1, | |
| "gradient_checkpointing": false, | |
| "model_type": "bart", | |
| "normalize_before": false, | |
| "normalize_embedding": true, | |
| "task_specific_params": { | |
| "summarization": { | |
| "length_penalty": 1.0, | |
| "max_length": 128, | |
| "min_length": 12, | |
| "num_beams": 4 | |
| }, | |
| "summarization_cnn": { | |
| "length_penalty": 2.0, | |
| "max_length": 142, | |
| "min_length": 56, | |
| "num_beams": 4 | |
| }, | |
| "summarization_xsum": { | |
| "length_penalty": 1.0, | |
| "max_length": 62, | |
| "min_length": 11, | |
| "num_beams": 6 | |
| } | |
| }, | |
| "output_attentions": false | |
| }, | |
| "covers": { | |
| "facebook/bart-base": "aadd2ab0ae0c8268c7c9693540e9904811f36177" | |
| }, | |
| "built_from": { | |
| "repo": "facebook/bart-base", | |
| "revision": "aadd2ab0ae0c8268c7c9693540e9904811f36177" | |
| }, | |
| "seed": 17, | |
| "dtype": "float32", | |
| "builder": "example/build_fixtures.py", | |
| "model_class": "BartForConditionalGeneration", | |
| "parameters": 3515968, | |
| "missing_parameters": [], | |
| "checkpoint_inspection": "parameter headers/indexes; bounded binary fallback", | |
| "numeric_health_scope": "CPU float32/bfloat16/float16; autocast for low-precision losses", | |
| "captured": { | |
| "call": "training step", | |
| "inputs": [ | |
| "attention_mask", | |
| "input_ids", | |
| "labels" | |
| ], | |
| "class": "BartForConditionalGeneration" | |
| }, | |
| "weights": "overfit", | |
| "gate": { | |
| "random": { | |
| "pass": false, | |
| "accepted_insensitive": {}, | |
| "threshold": 0.001, | |
| "deltas": { | |
| "input_ids:roll_batch": 0.007265642285346985, | |
| "input_ids:reverse": 0.0008658245205879211, | |
| "attention_mask:all_ones": null | |
| }, | |
| "weakest": "input_ids:reverse", | |
| "no_perturbation_defined": [], | |
| "not_a_forward_parameter": [] | |
| }, | |
| "overfit": { | |
| "pass": true, | |
| "accepted_insensitive": {}, | |
| "threshold": 0.001, | |
| "deltas": { | |
| "input_ids:roll_batch": 7.218916416168213, | |
| "input_ids:reverse": 0.6016836166381836, | |
| "attention_mask:all_ones": null | |
| }, | |
| "weakest": "input_ids:reverse", | |
| "no_perturbation_defined": [], | |
| "not_a_forward_parameter": [] | |
| } | |
| }, | |
| "training": { | |
| "seconds": 0.66, | |
| "steps": 60, | |
| "loss": 0.016765477135777473, | |
| "converged": true | |
| }, | |
| "numeric_health": { | |
| "float32": { | |
| "outputs_finite": true, | |
| "loss_finite": true, | |
| "gradients_finite": true, | |
| "finite": true | |
| }, | |
| "bfloat16": { | |
| "outputs_finite": true, | |
| "loss_finite": true, | |
| "gradients_finite": true, | |
| "finite": true | |
| }, | |
| "float16": { | |
| "outputs_finite": true, | |
| "loss_finite": true, | |
| "unscaled_gradients_finite": true, | |
| "finite": true | |
| } | |
| }, | |
| "copied_unchanged": [ | |
| "merges.txt", | |
| "tokenizer.json", | |
| "vocab.json" | |
| ], | |
| "environment": { | |
| "python": "3.12.12", | |
| "machine": "arm64", | |
| "torch": "2.14.1", | |
| "transformers": "5.22.0.dev0", | |
| "tokenizers": "0.23.3" | |
| }, | |
| "sha256": { | |
| "config.json": "06031d4f33259ccd79aa7d4a04aae90e2ac36b6a289c92e902fec85fd8c97451", | |
| "generation_config.json": "14ad8633f91c6394eaa6131a082e55f3d08d2aedd7840d2e3350a3668985d115", | |
| "merges.txt": "1ce1664773c50f3e0cc8842619a93edc4624525b728b188a9e0be33b7726adc5", | |
| "model.safetensors": "1016d81a693831a4ddb53d90c2ae1da51f9d457995df17c89e085bca66b8da5a", | |
| "tokenizer.json": "847bbeab6174d66a88898f729d52fa8d355fafe1bea101cf960dd404581df70e", | |
| "vocab.json": "9e7f63c2d15d666b52e21d250d2e513b87c9b713cfa6987a82ed89e5e6e50655" | |
| } | |
| } | |